1 paper · 1 filter
Daiki Yotsufuji, Kenta Nishihara, Shoma Shimizu +2
In challenging real-world reinforcement learning applications, communication delays or sensor failures often cause frame dropping, in which the agent cannot receive the dropped sta…